Professional Documents
Culture Documents
9 February 2011
File: AnnualGDP10a
Abstract: This paper provides the first annual GDP series for Great Britain over the
period 1700-1870, constructed in real terms from the output side. Sectoral estimates
are provided for agriculture, industry and services, and for a number of sub-sectors.
Estimates of nominal GDP are also provided, based on a benchmark for 1841 and
projected back to 1700 and forward to 1870 using the real output series and sectoral
price indices. The new data are used to provide a consistent account of economic
growth and the business cycle. The results are broadly consistent with the long run
path of real output suggested by Crafts and Harley, although growth rates for sub-
periods differ, largely as a result of changes in the growth of agriculture. Nominal
GDP increased more rapidly than suggested by Lindert and Williamson during the
eighteenth century, and more slowly than suggested by Deane and Cole during the
first half of the nineteenth century, as a result of differences in the price indices. We
also refine the business cycle chronologies of Ashton and Gayer, Rostow and
Schwartz.
Acknowedgements: This paper forms part of the project “Reconstructing the National
Income of Britain and Holland, c.1270/1500 to 1850”, funded by the Leverhulme
Trust, Reference Number F/00215AR. It is also part of the Collaborative Project HI-
POD supported by the European Commission's 7th Framework Programme for
Research, Contract Number SSH7-CT-2008-225342. A pilot study was conducted
with the financial support of the European Commission’s 6th Framework Programme
via the Research Training Network “Unifying the European Experience: Historical
Lessons of Pan-European Development”, contract no. MRTN-CT-2004-512439,
during which Péter Földvári provided excellent research assistance. We are grateful to
Bob Allen, Nick Crafts, Solomos Solomou and Jan Luiten van Zanden for helpful
comments and advice.
2
I. INTRODUCTION
This paper provides an annual series of gross domestic product for Great Britain over
the period 1700-1870, built up from the output side. For the period 1700-1830, the
estimates of Crafts and Harley (1992) provide observations of sectoral and aggregate
output only for a small number of benchmark years, while for the period 1830-1870,
Sectoral estimates of real GDP are provided here for agriculture, industry and
services, and also for a number of sub-sectors. Estimates of nominal GDP are also
constructed by combining the real output series with price indices to project
We use the new annual series to provide a consistent account of growth and
the business cycle in Britain over this period. The results are broadly consistent with
the long run path of real output suggested by Crafts and Harley (1992), although
growth rates for sub-periods differ, largely as a result of changes in the growth of
agriculture. Nominal GDP increased more rapidly than suggested by Lindert and
Williamson (1982) during the eighteenth century, and more slowly than suggested by
Deane and Cole (1967) during the first half of the nineteenth century, as a result of
differences in the price indices. We also refine the business cycle chronologies of
Ashton (1959) and Gayer, Rostow and Schwartz (1953). A sister paper by Broadberry
et al (2010) provides annual estimates of GDP for England covering the period 1270-
1700. The current paper thus forms a convenient link between this account of
medieval and early modern economic growth and the more familiar accounts of
British economic growth since 1870, based on the annual data of Feinstein (1972).
3
benchmark years, we demonstrate broad consistency of long run growth rates with the
widely-accepted approach of Crafts and Harley (1992). For each of the main sectors,
annual observations are graphed, while more information is also provided on the main
estimates of GDP in current as well as constant prices. The constant price GDP series
Section VII conducts an analysis of the business cycle in aggregate GDP and
II. AGRICULTURE
Deane and Cole (1967) measured agricultural output by assuming constant corn
consumption per capita and making a small adjustment for foreign trade. This was
criticised by Crafts (1976), who regarded the assumption of constant per capita
consumption of corn as indefensible while real incomes were rising and the relative
price of corn was changing. Crafts (1985: 39-42) used a hybrid method to calculate
the growth of agricultural output over the period 1700-1830. For the period 1700-60,
he used the demand approach with an assumed income elasticity of demand of 0.7 and
national income growth. For the period 1760-1801, and also for 1801-31, he allowed
for relative price changes with an assumed price elasticity of demand of -0.8, and
4
output directly using volume and price data published in the Agrarian History of
England and Wales by Chartres (1985), Bowden (1985), Holderness (1989) and John
(1989). He also showed that for the period 1700-1850 as a whole, the volume method
benchmark years covering the longer time span of 1300-1850, but with a somewhat
slower output growth over the period 1700-1850 than suggested by Allen (1994).
agriculture on an annual basis, built up from quantity and price information for
individual products. Results are presented for individual products and for the arable
and pastoral sub-sectors, as well as for total agriculture. The data are taken largely
from the Modern Farm Accounts Database assembled by Turner et al. (2001) and the
Early Modern Probate Inventories Database of Overton, Whittle, Dean and Haan
(2004), supplemented with additional information from Clark (2004), and the
Agrarian History of England and Wales. For arable farming we have annual gross
yields for all products for most years, although there are significant gaps for rye as it
For the pastoral sector, the data are drawn from the Early Modern Probate
Inventories Database between 1700 and 1750. Animal stocking densities (the number
5
of animals per sown acre) are multiplied with the sown acreage to yield estimates of
the total numbers of animals in England. After 1750, benchmark estimates of the
numbers of animals from John (1989) and Allen (2005) are interpolated with data on
the sale of animals at Smithfield from Mitchell (1988: 708-709). Multiplying the
resulting numbers of animals with the percentages of animals producing and the
2. Arable farming
The starting point for any estimate of the output of the arable sector is the total area
under crop, which is set out in Table 1. For most benchmark years, the data differ
subsequent scholarship. Firm estimates of land use only became available in the
agricultural returns of 1871, which therefore provides the starting point for the series.
For 1830, the figures come from the tithe files and for 1800, 1750 and 1700 from
figures suggested by Prince’s (1989: 41) interpretation of the 1801 Crop Returns.
The next step is to estimate average grain yields per acre. Weighted national
average yields per acre, gross of tithe and seed can be obtained from the Early
Modern Probate Inventories Database and the Modern Farm Accounts Database. Each
dataset has been divided into seven regional groupings and separate chronologies
have been constructed for each region before being combined into a single weighted
master chronology for the country as a whole. Due to the discontinuous nature of
much of the data, the regional chronologies are derived using regression analysis with
dummy variables for each farm and for each year, as suggested by Clark (2004).
6
Grain yields gross of seed as well as tithe are shown in Figure 1 for wheat, rye, barley,
oats and pulses. From these gross yields it is necessary to subtract grain used as seed
to derive the net yields shown in Table 2. Net yields increased strongly for all grains,
with oats and wheat showing bigger increases than barley and rye.
In addition to making allowance for grain used as seed, the derivation of net
output in arable farming requires the subtraction of grain used to feed working
animals. For the first half of the eighteenth century, estimates of the numbers of
working animals per sown acre can be obtained from the Early Modern Probate
Inventories Database, which can be multiplied with the sown acreage to produce
century, oxen numbers are assumed to decline steadily to zero by 1870, while horses
are taken from Allen (1994: 102). Using assumptions about consumption of oats and
animal consumption, which are then subtracted from gross output to arrive at arable
output net of seed and animal consumption in Table 4. The output of wheat, the
principal bread grain, increased steadily, while the output of rye, an inferior bread
grain, declined in absolute as well as relative terms. Barley, the principal brewing
grain saw a rapid growth during the first half of the nineteenth century, while oats,
which came to be used increasingly as horse fodder, grew more erratically. The output
of potatoes increased dramatically throughout the period, while the output of pulses
3. Pastoral farming
7
The starting point for the estimation of pastoral farming output is the numbers of non-
working animals in Table 5. For the first half of the eighteenth century, these are
derived from the stocking densities in the Early Modern Probate Inventories Database.
For later years, estimates for benchmark years are derived from Allen (2005), John
(1989), Mitchell (1988) and Turner (1998), and interpolated using data on annual
sales at Smithfield and the Metropolitan Cattle Market from Mitchell (1988: 708) and
Perren (1975: 388). All animal types increased in number, leading to a substantial
The proportions of animals assumed to have been producing milk, meat and
wool are set out in Table 6. A high proportion of cows are assumed to have produced
milk and a high proportion of sheep to have yielded wool. Meat, however, was
produced only by those animals that were slaughtered. Slaughter rates for 1700 and
1850 are derived from Holderness (1989) and Clark (1991). The next step in the
calculation of pastoral output involves the estimation of yields of milk, meat and wool
per animal, set out in Table 7. Data between benchmark years were interpolated using
information on the relative prices of pastoral products and the animals from which
they were derived. Table 8 then combines the information on numbers of animals,
percentages producing and yields per animal to arrive at the estimates of total output
Further assumptions are needed to derive output estimates for hay, hides and
skins and dairy products. Hay output is derived from the numbers of non-farm horses,
on the assumption that each horse consumed 2.4 tons of hay per year (Allen, 2005).
Output of hides and skins is derived from the numbers of working and non-working
8
animals, using assumptions on the percentages of each animal producing and yields
per animal from Clark (1991) and Clarkson (1989). In the dairy sector, output is split
between cheese, butter and fresh milk using data from Holderness (1989).
Multiplying the output volumes by their prices yields the total value of net output.
The price data are taken largely from Clark (2004), who synthesises the published
data of Beveridge (1939), Thorold Rogers (1866-1902: volumes 1-30) and the multi-
volume Agrarian History of England and Wales, as well as integrating new archival
material. Output can be valued in both current prices and in constant 1700 prices.
Figure 2 plots arable, pastoral and total agricultural output in constant prices
on a logarithmic scale. Pastoral output grew at a faster rate than arable output,
particularly from the mid-eighteenth century. The pastoral sector thus increased its
share of agricultural output in constant prices. In Table 9, we see that the share of
pastoral products in current prices also increased substantially from the mid-
eighteenth century, as relative prices changed only over relatively short periods.
One finding from Table 9 that is worth emphasising is the high share of the
pastoral sector in agricultural value added, even at the beginning of the eighteenth
century. As a result of its subsequent growth, the pastoral sector accounted for more
than 60 per cent of agricultural output by the late nineteenth century. This meant that
although the British people did not have a particularly generous diet if viewed in
terms of kilocalories, it was a varied diet, with meat, dairy produce and ale to
supplement the less highly processed grain products that made up the bulk of the diet.
9
Table 10 presents the total agricultural output index from Figure 2 in growth
rate form, using both annual data and 10-year averages to capture long run trends.
Over the 1700-1830 period as a whole, our agricultural growth rate is very similar to
that of Crafts (1985) and Crafts and Harley (1992), although there are considerable
differences over shorter periods. In particular, in common with the volume indices
calculated by Allen (1994: 102) and Overton (1996: 75), we do not find the marked
slowdown in the third quarter of the eighteenth century suggested by Crafts and
Harley. Rather, we find the slowdown concentrated in the first quarter of the
nineteenth century.
III. INDUSTRY
Industry is the one sector for which data have previously been analysed at annual
(1985) and Harley (1982) pointed out independently, Hoffmann (1955) inadvertently
overstated the growth rate of industrial output during the Industrial Revolution as a
result of his weighting procedures. The problem is that a few industrial branches,
most notably cotton and iron, grew much more rapidly than the rest of industry, and
these branches are included in Hoffmann’s data set. However, the available time
series cover only 56 per cent of industrial output, and the weights of these industries
are increased proportionally to achieve 100 per cent coverage of industrial output. But
this means that the unrepresentative, rapidly growing branches of cotton and iron
effectively have their weights doubled. Harley (1982) and Crafts et al. (1989) propose
10
that only the weights of industries other than cotton and iron should be increased to
In addition to changing the weighting scheme, Harley (1982) and Crafts et al.
(1989) also replaced some of the older series used by Hoffmann (1955), drawing on
the latest scholarship. We use these series, together with some later additions, the
most important of which are the new series of bar iron output from King (2005),
Feinstein’s (1988: 446) series of investment in total buildings and works for output of
the building industry, and an index of new English language book titles derived from
the English Short Title Catalogue and the British Library for the output of the printing
industry. Our weighting scheme is very similar to that of Crafts et al. (1989), but
Figure 3 presents our series for industrial output, together with the “revised best
guess” series of Crafts and Harley (1992) and Hoffmann’s (1955) original index for
contrast. The biggest difference is between the Hoffmann index and the other two
indices, as a result of the excessive weight given to cotton textiles and iron in the
former. Our series shows slightly slower growth than the Crafts-Harley index during
the early eighteenth century, largely as a result of the inclusion of new series,
particularly King’s (2005) data for the iron industry. From the mid-eighteenth century
onwards, differences between the two series are relatively minor, and essentially
confirm the picture originally presented in Crafts et al. (1989). Output growth
accelerated from around 1740 to 1840 before tapering off. Again, there is quite a
growth rates of industrial output over the conventional sub-periods calculated using
both the raw annual data and 10-year averages, together with the Crafts-Harley
building and mining. Both mining and building grew more rapidly than total
manufacturing over the period as a whole. Part B of Figure 4, however, shows that
total manufacturing output included some very rapidly growing branches. The most
rapid growth was in metal production, driven by the iron industry. The next most
rapid growth was in textiles, driven by the dramatically expanding cotton industry, but
slowed down by the relative decline of the more traditional textile industries. Food
IV. SERVICES
The service sector has received much less attention from economic historians than
agriculture and industry, and there have been no previous attempts to provide annual
frequency data. The first estimates for benchmark years were provided by Deane and
Cole (1967), but have been revised downwards by Crafts (1985), particularly for the
early nineteenth century. For the eighteenth century Deane and Cole (1967: 76-78)
assumed that “commerce” grew at the same rate as industry, that “rent and
miscellaneous services” increased in line with population, and that “government and
defence” could be measured by real public expenditure. Crafts (1985: 35-37) made
only minor changes here. For the nineteenth century, however, Deane and Cole (1967:
12
166) derived estimates of income in “trade and transport”, “domestic and personal”,
“housing”, “government, professional and other services” and deflated them by the
Rousseaux price index. Crafts (1985: 31) showed that this produces an implausibly
high rate of growth for commerce, and assumed instead that commerce grew in line
with national income, thus introducing an element of iteration into the estimates.
Crafts (1985: 35-37) used employment growth for domestic and personal services and
new estimates of the housing stock for housing. He also revised the growth rate of
Our estimates are broadly consistent with those of Crafts (1985), but make a
few changes to reflect the need for annual data. The most important difference is in
commerce, where we measure the growth of output using volume series covering
distribution, transport, finance and other commerce. This produces results which are
not far out of line with the Deane and Cole (1967) assumption for the eighteenth
century, that commerce grew in line with industry. This also avoids the iterative
element in the Crafts (1985) assumption that commerce grew in line with national
income during 1801-30, and ensures consistency of treatment throughout the whole
period.
the whole period, deflated using the Schumpeter-Gilboy and Rousseaux price indices
from Mitchell (1988: 719-723). For housing, we use the stock estimates of Feinstein
(1988: 389), using a regression relationship between housing stock and population to
fill in gaps. Output of domestic and personal services is assumed to rise in line with
population throughout the whole period, following Deane and Cole (1967) and Crafts
13
(1985). This inevitably produces a relatively stable path for output in domestic and
Our annual index of service sector output is plotted in Figure 5. The trend pattern is of
an increase in the growth rate from around 1780. As is usually the case, fluctuations
in services were milder than in industry or agriculture. Table 13 presents the annual
growth rates of services output over the conventional sub-periods calculated using
both the raw annual data and 10-year averages, together with the Crafts-Harley
estimates for comparison. Our growth rate estimates are very close to the Crafts-
Harley data over both the whole period 1700-1830 and individual sub-periods, and
Figure 6 provides more detail on the service sector. Part A provides information on
commerce, government, housing and domestic services, while Part B breaks down
commerce into transport, distribution, finance and other commerce. In Part A, we see
that government grew most rapidly during the eighteenth century, albeit in a sharply
cyclical fashion as a result of warfare. During the nineteenth century, commerce was
clearly the fastest growing sector. Housing and domestic services grew more slowly
but more steadily than government or commerce. The former two series coincide in
the chart because domestic services were assumed to grow in line with population,
whilst we found a unit elasticity between housing and population. In Part B, we see
that distribution and finance grew more rapidly than transport and other commerce,
14
with finance particularly dynamic during the eighteenth century, but distribution more
The next step is to construct an index of real GDP from the above output series for
agriculture, industry and services, using an appropriate set of weights. Table 14 sets
out the weighting scheme, derived from the reconstruction of nominal GDP by sector.
Real output trends from the sectoral series described earlier in the paper are
transformed into current price trends using sectoral price deflators. For agriculture we
construct a price index based on a weighted average of price series for individual
agricultural products, taken largely from Clark (2004). The price index for the
industrial sector is based on a weighted average of price series for industrial products
from Clark (2004) and Beveridge (1939). For service sector prices, the key data are
wage rates and house rents from Clark (2004), with some limited information on
transport prices from Harley (1988) and Bogart (2005), and with distribution prices
table for 1841, derived from Horrell et al. (1994), but adjusted from a United
Kingdom to a Great Britain basis using the information in Horrell et al. (1991).
Aggregate real GDP is constructed from the sectoral real output series using current
price sectoral shares as follows: 1700 weights are used for the period 1700-1740,
1759 weights for 1740-1780, 1801 weights for 1780-1820 and 1841 weights for 1820-
1870.
15
Putting the three main sectors together using the weights from Table 14, we
arrive at the annual index of real GDP shown in Figure 7 and presented in growth rate
form in Table 15, together with the Crafts-Harley data for comparison. Our series
shows much the same pattern of trend growth acceleration as the Crafts-Harley data,
although there are some minor differences over shorter sub-periods, largely as a result
of differences in agriculture. Figure 7 shows clearly that the fastest growth was in
Using the price series for agriculture, industry and services to reflate real
output in each sector yields the values for current price GDP shown in Figure 8. For
comparison, we also show Deane and Cole’s (1967) current price GDP estimates for
benchmark years in the nineteenth century, together with Lindert and Williamson’s
(1982) benchmark figures for the eighteenth century. There is broad agreement over
the long run trends, but our estimates suggest a higher nominal GDP in most years. A
closer examination of the data for benchmark years in Table 16 reveals that during the
nineteenth century our estimates of nominal GDP grow substantially more slowly
than the Deane and Cole (1967) estimates between 1801 and 1851, but more rapidly
after 1851. For the eighteenth century, our estimates grow substantially more rapidly
Figure 9 plots the price indices for the three main sectors together with the
aggregate price index. Agricultural and service sector prices increased substantially
during the second half of the eighteenth century, rising to a peak at the end of the
Napoleonic Wars before falling back during the postwar deflation. By contrast,
16
industrial prices increased much less during the eighteenth century and fell
substantially more after 1815. Figure 10 therefore shows the expected pattern for the
inter-sectoral terms of trade during the Industrial Revolution, with the relative price of
Figure 11 plots real and nominal GDP together with the aggregate price level
over the period 1700-1870. Most of the increase in nominal GDP was the result of
real growth, with only a modest increase in the price level. Although there was a
period of substantial inflation during the French and Napoleonic Wars, this was
the total population of Great Britain. From 1801 onwards, annual data on the
population of England, Wales and Scotland are available from Mitchell (1988: 9). For
the period before 1801, the population of England has been reconstructed firmly by
Wrigley and Schofield (1989) and Wrigley et al. (1997) for the period since the
available for Wales and Scotland, we assume that the ratio of the population of Wales
to England remained the same for the period 1700-1801. For Scotland, we have
population estimates for 1700 and 1750 (Schofield 1994: 93). Other years are
Combining the GDP series with the population data produces our estimates of
per capita income in Figure 12. Table 17 presents the same material in growth rate
17
form. The main findings are that per capita income growth accelerated considerably
between 1780 and 1801, and then slowed down between 1801 and 1830, before
accelerating again after 1830. For the period 1700-1870 as a whole, per capita income
grew at an annual rate of 0.48 per cent using the 10-year average data.
The conventional business cycle chronology for the period 1700-1870 was established
by Ashton (1959) for the eighteenth century and by Gayer et al. (1953) and Rostow
(1972) for the nineteenth century, based on the NBER methodology of Burns and
Mitchell (1946). This involved checking a large number of microeconomic time series
statistical data rather than turning-points in any particular magnitude such as national
income” (Matthews, 1954: 2). The peak and trough years obtained with this
methodology are shown in Table 18. The minor disagreements over turning points in
the overlap period can be put down to the difficulties of assigning tuning points on an
annual basis when monthly data would be more appropriate. Since growth rates are
some bias in the conventional periodisation used in the analysis of British economic
growth. Picking peak years on the basis of the NBER methodology, the appropriate
periodisation would be 1701, 1761, 1783, 1802, 1831 and 1866 rather than 1700,
Table 19 sets out the business cycle chronology obtained from the real GDP
series constructed above, using the Hodrick-Prescott filter, with the smoothing
18
parameter λ set equal to 100, following Backus and Kehoe (1992).1 Using a lower
value of λ equal to 6.25, as suggested by Ravn and Uhlig (2002), reduces the
amplitude but not the frequency of the business cycle. The cycle with λ equal to 100
is graphed in Figure 13. If GDP can be taken as a more appropriate summary measure
of “general business activity”, the correct periodisation would be 1701, 1761, 1781,
1802, 1830, 1868. Given the disagreements on the precise dates of the turning points,
we have taken 10-year averages based on the years traditionally analysed in the
growth literature.
VIII. CONCLUSIONS
This paper provides the first estimates of GDP on an annual basis for Great Britain
during the period 1700-1870. This enables us to analyse economic growth and the
business cycle within a consistent framework. The results are broadly consistent with
the long run path of real output suggested by Crafts and Harley (1992), although
growth rates for sub-periods differ, largely as a result of changes in the growth of
agriculture. Nominal GDP increased more rapidly than suggested by Lindert and
Williamson (1982) during the eighteenth century, and more slowly than suggested by
Deane and Cole (1967) during the first half of the nineteenth century, as a result of
differences in the price indices. We also refine the business cycle chronologies of
1
See Hodrick and Prescott (1997).
19
Wheat Rye/ Barley/ Oats Pulses Potatoes Other Total Fallow Total
Maslin Dredge crops sown arable arable
1700 1.99 0.42 1.82 1.15 0.98 0.00 1.30 7.64 1.91 9.56
1750 1.95 0.06 1.50 1.82 0.98 0.08 2.53 8.92 1.59 10.51
1800 2.51 0.06 1.46 1.97 0.83 0.17 2.90 9.91 1.28 11.19
1830 2.12 0.06 1.81 1.27 0.63 0.26 4.46 10.62 1.30 11.91
1871 3.31 0.06 1.96 1.45 0.90 0.39 5.28 13.35 0.48 13.84
Sources: Overton and Campbell (1996: Tables III, V); Early Modern Probate
Inventory Database; Holderness (1989); Overton (1996).
FIGURE 1: English weighted national average grain yields per acre, gross of
tithe and seed (bushels, log scale)
A. Wheat
20
FIGURE 1 (continued): English weighted national average grain yields per acre,
gross of tithe and seed (bushels, log scale)
B. Rye
C. Barley
21
FIGURE 1 (continued): English weighted national average grain yields per acre,
gross of tithe and seed (bushels, log scale)
D. Oats
Sources: Early Modern Probate Inventories Database and Modern Farm Accounts
Database.
22
TABLE 2: English mean yields per acre gross of tithes, net of seeds in bushels
(10-year averages)
Sources and notes: Gross yield per acre taken from the Early Modern Probate
Inventories Database and the Modern Farm Accounts Database. Seed sown per acre
from the Modern Database. Pulses are taken from Overton and Campbell (1996),
Allen (2005).
Oxen Horses
1700-1709 0.10 0.57
1750-1759 0.07 0.78
1800-1809 0.03 0.89
1850-1859 0.00 1.25
1861-1870 0.00 1.26
Sources: Derived from the Early Modern Probate Inventories Database; Allen (1994);
John (1989); Turner (1998).
TABLE 4: English arable output net of seed and animal consumption in million
bushels (10-year averages)
Source: Output gross of tithe and net of seed were derived by multiplying sown area
from Table 1 with net yields from Table 2. The sown area from Table 1 was
interpolated where necessary. Consumption by working animals was derived from the
numbers of working animals shown in Table 3. For oats, outlying observations based
on a very small number of inventories were dropped in 1700-09 and 1750-59, to
eliminate excessive volatility.
23
Sources and notes: Derived from Early Modern Probate Inventory Database; Allen
(2005); John (1989 Tales III.1 and III.2), Mitchell (1988); Turner (1998).
* Livestock units compare different animals on the basis of relative feed
requirements. Ratios from Campbell (2000: 104-107): (adult cattle for beef and milk x
1.2) + (immature cattle x 0.8) + (sheep and swine x 0.1).
Sources and notes: Milk, beef, mutton, pork, and wool are obtained from Clark (1991:
216), while veal is taken from Allen (2005: Table 6). The missing years were
interpolated in line with the ratio of product to animal prices.
24
Sources: Total output estimates are derived by multiplying animal numbers from
Table 5 with the percentage of animals producing in Table 6. The resulting numbers
of producing animals are then multiplied with the animal yields from Table 7.
FIGURE 2: Indexed output in English arable and pastoral agriculture (log scale,
1700=100)
A. Arable products
Year Wheat Rye Barley Oats Pulses Potatoes Total arable
products
1700-09 27.3 4.6 16.8 0.1 5.1 0.4 54.4
1750-59 26.4 0.6 12.8 6.8 3.9 6.0 56.6
1800-09 24.0 0.4 10.6 4.3 3.5 3.4 46.2
1850-59 21.1 0.2 10.5 2.5 1.9 6.7 42.9
1861-70 19.3 0.1 9.0 1.9 1.9 6.1 38.2
B. Pastoral products
Year Total
pastoral
Dairy Beef Pork Mutton Hay Wool Hides products
1700-09 12.0 2.8 4.4 15.5 4.0 5.6 1.2 45.6
1750-59 12.3 3.3 4.8 13.5 5.0 3.5 1.2 43.4
1800-09 15.9 5.2 5.7 15.4 7.7 3.0 0.9 53.8
1850-59 17.8 6.3 7.2 15.0 6.8 3.2 0.8 57.1
1861-70 20.0 6.9 7.7 16.2 6.0 4.1 0.8 61.8
Sources: Derived from Early Modern Probate Inventories Database; Modern Farm
Accounts Database.
Sources: Crafts (1985: 45; Crafts and Harley (1992: 715); see text.
26
Sources: Crafts (1985: 32); Crafts and Harley (1992: 715); see text.
27
FIGURE 5: British service sector output in real terms, 1700-1870 (log scale,
1700=100)
Sources: Derived from Crafts (1985: 16-17, 32, 37); Crafts and Harley (1992: 715);
see text.
29
Sources and notes: Derived from reconstruction of nominal GDP by sector. Real
output trends above are transformed into current price trends using sectoral price
deflators, with absolute levels of GDP in current prices established using an input-
output table for 1841, based on Horrell et al. (1994). 1700 weights are used for the
period 1700-1740, 1759 weights for 1740-1780, 1801 weights for 1780-1820 and
1841 weights for 1820-1870.
Sources: Crafts (1985: 45); Crafts and Harley (1992: 715); see text.
Sources: Deane and Cole (1967); Lindert and Williamson (1982); see text.
32
Sources and notes: See text; Deane and Cole (1967); Lindert and Williamson’s (1982)
estimates for England converted to a Great Britain basis using data on the population
share of Scotland and Scottish per capita incomes as a percentage of the average for
Great Britain from income tax data in Lee (1986: 127, 131).
FIGURE 11: Real and nominal GDP, Great Britain 1700-1870 (1700=100, log
scale)
FIGURE 12: British real GDP per capita, 1700-1870 (log scale, 1700 = 100)
TABLE 17: Average annual growth rate of British population and per capita
income, 1700-1870 (% per annum)
Sources: Mitchell (1988), Wrigley and Schofield (1989), Schofield (1994) and
Wrigley et al. (1997); see text.
35
A. Eighteenth Century
Peak Trough Peak Trough
1701 1702 1751 1755
1704 1706 1761 1763
1708 1712 1764 1769
1714 1716 1771-72 1775
1717-18 1722 1777 1781
1724-25 1727 1783 1784
1728 1730 1787 1789
1733 1734 1792 1794
1738 1742 1796 1798
1743 1746 1799 1800
1746 1748 1802
B. Nineteenth Century
Peak Trough Peak Trough
1792 1793 1828 1829
1796 1797 1831 1832
1800 1801 1836 1837
1802 1803 1839 1842
1806 1808 1845 1848
1810 1811 1854 1855
1815 1816 1857 1858
1818 1819 1860 1862
1825 1826 1866 1868
Sources: Ashton (1959: 172); Gayer et al. (1953: 348); Rostow (1972: 77).
36
FIGURE 13: Cyclical component of the log of the GDP index (1850=100) after
Hodrick-Prescott filter (λ=100)
2. Industry
The weighting scheme presented in Table A1 is based on Hoffmann (1955), but
modified as discussed in the text. The basic data are obtained from Hoffmann (1955),
apart from the modifications listed below:
Mining: coal: Pollard’s (1980) decadal estimates, interpolated using the Hoffmann
series.
Metals: Iron, steel and machine building: Hoffmann’s series is replaced for the period
before 1839 using bar iron output from King (2005). The yearly data can be
downloaded from the Economic History Society website at
http://www.ehs.org.uk/ehs/Datasets/datasets.asp. Shipbuilding: Gross capital
formation in ships from Feinstein (1988: 446), interpolated using the Hoffmann
series.
Textiles: Silk thread and goods: Mitchell (1988: 343) reports the imports of raw,
thrown and waste silk for 1700-1825. These are used to extend the Hoffmann series.
Linen yarn and cloth: Hoffmann’s series extended using linen yarn imports from
Schumpeter (1960: 52).
Food, Drink & Tobacco: Beer: Prior to 1787, the data are brought back in time by
connecting an index of small, strong, and table beer charged with duty (Mitchell
1988: 404-405) to the Hoffmann index. Tobacco products: This series is extended
back to 1700 with the value of imported tobacco at official prices from Mitchell
(1988: 462-463).
Other Manufacturing: Paper: Paper charged with duty from Mitchell (1988: 413).
Printed matter: Index of new English language book titles (1700-1800) from the
English Short Title Catalogue (http://www.rlg.org/estc.html). This is extended from
1800 to 1870 with an index of new titles from the Integrated Catalogue of the British
Library (http://catalogue.bl.uk/F/?func=file&file_name=login-bl-list). The
overlapping years show a high correlation, thus these two series can be linked
together. Leather: Hoffmann’s series is extended back from 1801 to 1722 with an
index of hides and skins charged with duty from Mitchell (1988: 416). Soap and
candles: For the period prior to 1821, separate series are available for soap and
candles from Mitchell (1988: 412, 415). Soap charged with duty is available back to
1713, while tallow candles charged with duty are available back to 1711.
Building: The Hoffmann series is replaced with a constant price series of total
buildings and works from Feinstein (1988: 446), available for 1761-1870 but with
gaps. Timber imports from Mitchell (1988: 462) were used to interpolate.
3. Services
Weights of the four main service sectors are given as shares of total services in Table
A2. These shares are taken from Crafts (1985: 16-17). Data sources are as follows:
3.1. Commerce
Transport: We used shipping tonnage from Mitchell (1988: 534), the length of
railways from Mitchell (1988: 541), total investment in waterways and roads from
Ginarlis and Pollard (1988: 217-219) and the number of weekly passenger road
39
services from Bogart (2005: 487), which we interpolated assuming constant rates of
growth. The index is created from the unweighted average of the available series.
Financial services: We used the number of country banks from Pressnell (1956: 11),
available for 1784, 1793, 1796, 1798, 1800-1842. These points were interpolated
using the drawing accounts of the Bank of England from Mitchell (1988: 665) for
1720-1844, extended to 1870 with total deposits from Mitchell (1988: 658). The
period 1700-1720 was projected using an exponential trend. We also used Pearson’s
(2004: 374-375) series on fire insurance (real sum insured). The composite index was
obtained as the unweighted average of the banking and insurance indices.
Distribution: In the spirit of Deane and Cole’s (1967) work on the eighteenth century,
we assumed that the growth of distribution is a weighted average of the growth of
foreign trade and industry, with weights of 0.6 and 0.4 respectively. Checking the
sensitivity of the results to changing the weights, we found that reducing the share of
foreign trade to 30 per cent reduced the growth rate over the period 1700-1870 by just
0.2 percentage points. The data on exports and imports are taken from Mitchell (1988:
448-450).
Other commerce: We assume that other commerce grew in line with industry.
3.3 Government
We use civil government and defence expenditure from Mitchell (1988: 578, 587),
deflated using the Schumpeter-Gilboy price index to 1800 and the Rousseaux price
index after 1800, both from Mitchell (1988: 719, 722).
3.4 Housing
We use housing stock data from Feinstein (1988: 389), using a regression relationship
between housing stock and population to fill in gaps.
B. PRICE SERIES
1. Agriculture
Data are from Clark (2004, 2005, 2007), and can be downloaded from:
http://www.iisg.nl/hpw/data.php#united.
2 Industry
Unless otherwise specified, data are from Clark (2004, 2005, 2007), and can be
downloaded from: http://www.iisg.nl/hpw/data.php#united. Commodities used are as
follows:
Textiles: cotton, cotton cloth, wool, wool cloth, silk thread, linen cloth.
Metals: iron manufactures, pewter.
Other manufacturing: wood, paper-foolscap, books, leather (Beveridge, 1939; the
average of Naval Stores leather, backs and hose), soap, candles, lamp oil, coal gas.
Construction: bricks, tiles (Beveridge, 1939; the average of Winchester, Eton,
Westminster, Sandwich, Greenwich, Office of Works, Naval Stores. We used paving,
roof, and plain tiles but omitted ridge and paving tiles from the Naval Stores because
these were outliers compared to the other series), laths (Beveridge, 1939; the average
of Greenwich and Office of Works), lime (Beveridge, 1939; the average of
Winchester, Eton, Westminster, Sandwich, Greenwich, Office of Works, Naval
40
Stores), sand (Beveridge, 1939; the average of Westminster, Office of Works, Naval
Stores; we used masons’ sand and gravel, and sand, cement and tarras (Beveridge
1939; the average of cement, mortar, and tarras from Greenwich and Westminster),
lead (Beveridge, 1939; the average of lead, milled sheet, sheet and cast, pipe from
Westminster, Greenwich, Office of Works, and Naval Stores), building labourer’s
wage.
Mining: coal.
Foodstuff: wheaten flour, bread, bacon, treacle, sugar, beer, malt (Beveridge, 1939;
the average of Winchester, Eton, Westminster, Greenwich and Navy Victualling in
London, Portsmouth, and Plymouth), spirits, tobacco.
3 Services
Unless otherwise specified, data are from Clark (2004, 2005, 2007), and can be
downloaded from: http://www.iisg.nl/hpw/data.php#united. Series used are as
follows:
Housing: rent.
Domestic service: wages of building labourers.
Government: wages of craftsmen.
Financial services: fire insurance (Pearson, 2004: 374-380)
Transport: comprised of shipping (Harley, 1988; 873-875); goods road transport
(Bogart 2005: 505, interpolated decadal figures, 1700-1830); passenger road transport
(Bogart 2005: 506, interpolated decadal figures, 1750-1830).
Distribution: weighted average of agriculture and industry prices.
41
1700-1711 Coal, 11.4; Iron, steel and machine building, 11.8; Tin, 1.7; Cotton
yarn and cloth, 8.8; Silk thread and goods, 11.4; Linen yarn and cloth,
21.2; Sugar, 0.8; Beer, 14.0; Malt, 4.4; Tobacco products, 2.1; Printed
matter, 3.6; Building, 8.7
1711-1713 Coal, 10.7; Iron, steel and machine building, 11.8; Tin, 1.6; Cotton
yarn and cloth, 8.2; Silk thread and goods, 10.7; Linen yarn and cloth,
19.7; Sugar, 0.7; Beer, 13.0; Malt, 4.1; Tobacco products, 2.0; Printed
matter, 3.6; Candles, 5.2; Building, 8.7
1713-1722 Coal, 11.3; Iron, steel and machine building, 13.2; Tin, 1.4; Cotton
yarn and cloth, 7.3; Silk thread and goods, 9.5; Linen yarn and cloth,
17.7; Sugar, 0.6; Beer, 11.7; Malt, 3.6; Tobacco products, 1.8; Paper,
0.8; Printed matter, 4.0; Soap, 2.6; Candles, 4.7; Building, 9.7
1722-1727 Coal, 5.9; Iron, steel and machine building, 8.2; Tin, 0.8; Cotton yarn
and cloth, 4.6; Silk thread and goods, 5.9; Linen yarn and cloth, 10.8;
Sugar, 0.3; Beer, 7.1; Malt, 2.2; Tobacco products, 1.0; Paper, 0.5;
Printed matter, 4.1; Leather and leather goods, 34.1; Soap, 1.6;
Candles, 2.9; Building, 9.9
1727-1739 Coal, 5.8; Copper ore, 0.5; Iron, steel and machine building, 8.1; Tin,
0.8; Cotton yarn, 4.5; Silk thread and goods, 5.8; Linen yarn and
cloth, 10.7; Sugar, 0.3; Beer, 7.1; Malt, 2.2; Tobacco products, 1.0;
Paper, 0.5; Printed matter, 4.1; Leather and leather goods, 33.9; Soap,
1.6; Candles, 2.9; Building, 9.9
1739-1761 Coal, 4.0; Copper ore, 0.4; Iron, steel and machine building, 5.6; Tin,
0.5; Cotton yarn and cloth, 2.4; Woollen and worsted yarn and cloth,
27.5; Silk thread and goods, 4.0; Linen yarn and cloth, 7.4; Sugar, 0.3;
Beer, 5.0; Malt, 1.5; Tobacco products, 0.8; Paper, 0.3; Printed matter,
4.1; Leather and leather goods, 23.2; Soap, 1.1; Candles, 2.0;
Building, 9.9
1761-1771 Coal, 3.9; Copper ore, 0.4; Iron, steel and machine building, 6.5; Tin,
0.5; Cotton yarn and cloth, 6.7; Woollen and worsted yarn and cloth,
27.1; Silk thread and goods, 3.9; Linen yarn and cloth, 7.3; Sugar, 0.3;
Beer, 4.9; Malt, 1.5; Tobacco products, 0.8; Paper, 0.3; Printed matter,
3.4; Leather and leather goods, 19.7; Soap, 1.1; Candles, 2.0;
Building, 9.8
1771-1780 Coal, 3.8; Copper ore, 0.3; Iron, steel and machine building, 6.5;
Copper, 0.4; Tin, 0.5; Cotton yarn and cloth, 6.7; Woollen and
worsted yarn and cloth, 27.0; Silk thread and goods, 3.9; Linen yarn
and cloth, 7.2; Sugar, 0.3; Beer, 4.9; Malt, 1.5; Tobacco products, 0.8;
Paper, 0.3; Printed matter, 3.4; Leather and leather goods, 19.6; Soap,
1.1; Candles, 1.9; Building, 9.8
1780-1787 Coal, 3.8; Copper ore, 0.3; Iron, steel and machine building, 6.5;
Copper, 0.4; Tin, 0.5; Cotton yarn and cloth, 6.7; Woollen and
worsted yarn, 12.2; Woollen and worsted cloth, 14.8; Silk thread and
goods, 3.9; Linen yarn and cloth, 7.2; Sugar, 0.3; Beer, 4.9; Malt, 1.5;
Tobacco products, 0.8; Paper, 0.3; Printed matter, 3.4; Leather and
leather goods, 19.6; Soap, 1.1; Candles, 1.9; Building, 9.8
42
1787-1789 Coal, 3.8; Copper ore, 0.3; Iron, steel and machine building, 6.5;
Copper, 0.4; Tin, 0.5; Cotton yarn and cloth, 6.7; Woollen and
worsted yarn, 12.2; Woollen and worsted cloth, 14.7; Silk thread, 1.3;
Silk goods, 3.0; Linen yarn and cloth, 7.2; Sugar, 0.3; Beer, 4.9; Malt,
1.5; Tobacco products, 0.8; Paper, 0.3; Printed matter, 3.4; Leather
and leather goods, 19.5; Soap, 1.1; Candles, 1.9; Building, 9.7
1789-1801 Coal, 3.7; Copper ore, 0.3; Iron, steel and machine building, 6.5;
Copper, 0.4; Tin, 0.5; Shipbuilding, 1.9; Cotton yarn and cloth, 6.7;
Woollen and worsted yarn, 11.9; Woollen and worsted cloth, 14.4;
Silk thread, 1.2; Silk goods, 2.5; Linen yarn and cloth, 7.0; Sugar, 0.3;
Beer, 4.8; Malt, 1.4; Tobacco products, 0.8; Paper, 0.3; Printed matter,
3.4; Leather and leather goods, 19.1; Soap, 1.1; Candles, 1.9;
Building, 9.8
1801-1831 Coal, 8.6; Copper ore, 1.0; Iron, steel and machine building, 11.5;
Copper, 0.9; Copper products, 0.9; Tin, 0.4; Shipbuilding, 2.6; Cotton
yarn, 5.1; Cotton cloth, 10.1; Woollen and worsted yarn, 6.9; Woollen
and worsted cloth, 6.9; Silk thread, 0.6; Silk goods, 1.4; Linen yarn
and cloth, 5.3; Wheaten flour, 1.4; Bread and cakes, 3.0; Sugar, 0.5;
Beer, 0.9; Malt, 0.6; Spirits, 1.4; Tobacco products, 0.6; Paper, 1.9;
Printed matter, 3.9; Leather, 1.3; Leather goods, 8.8; Soap and
candles, 2.0; Building, 11.5
1831-1850 Coal, 10.7; Tin ore, 0.3; Copper ore, 0.8; Lead ore, 0.6; Iron, steel and
machine building, 12.2; Copper, 0.6; Copper products, 1.0; Lead, 0.3;
Tin, 0.1; Shipbuilding, 1.4; Furniture, 2.6; Timber products, 4.3;
Cotton yarn, 11.3; Cotton cloth, 6.3; Woollen and worsted yarn, 3.8;
Woollen and worsted cloth, 4.1; Silk thread, 1.0; Silk goods, 2.3;
Linen yarn and cloth, 2.5; Hemp products, 0.1; Wheaten flour, 2.4;
Bread and cakes, 1.6; Confectionary, 0.4; Sugar, 0.4; Beer, 2.3; Malt,
0.4; Spirits, 0.6; Tobacco products, 0.5; Paper, 1.5; Printed matter,
3.1; Leather, 1.1; Leather goods, 7.0; Soap and candles, 1.0;
Vegetable oils, 0.1; Building, 11.4
REFERENCES
Allen, R.C. (1994), “Agriculture during the Industrial Revolution”, in Floud, R. and
McClosky, D.N. (eds.), The Economic History of Britain since 1700, Second
Edition: Volume 1: 1700-1860, Cambridge: Cambridge University Press, 96-
122.
Allen, R.C. (2005), “English and Welsh Agriculture, 1300-1850: Output, Inputs, and
Income”, Oxford.
Backus, D.K. and Kehoe, P.J. (1992), “International Evidence on the Historical
Properties of Businesss Cycles”, American Economic Review, 82, 864-888.
Beveridge, W.H. (1939), Prices and Wages in England from the Twelfth to the
Nineteenth Century, Vol. I, Price Tables: Mercantile Era, London: Longmans,
Green.
Bogart, D. (2005), “Turnpike Trusts and the Transport Revolution in 18th century
England”, Explorations in Economic History, 42, 479-508.
Bowden, P.J. (1985), “Appendix III: Statistics”, in Thirsk, J. (ed.), The Agrarian
History of England and Wales, Volume V.II, 1640-1750: Agrarian Change,
Cambridge: Cambridge University Press, 827-902.
Broadberry, S.N., Campbell, B.M.S., Klein, A., Overton, M. and van Leeuwen, B.
(2010), “British Economic Growth, 1270-1870”, University of Warwick,
http://www2.warwick.ac.uk/fac/soc/economics/staff/faculty/broadberry/wp.
Broadberry, S.N., Campbell, B.M.S. and van Leeuween, B. (2010), “The Sectoral
Distribution of the Labour Force and Labour productivity in Britain, 1381-
1851”, University of Warwick.
Burns, A.F. and Mitchell, W.C. (1946), Measuring Business Cycles, New York:
National Bureau of Economic Research.
Clark, G. (2005), “The Condition of the Working Class in England, 1209 to 2004”,
Journal of Political Economy, 113, 1307-1340.
Clark, G. (2007), “The Long March of History: Farm Wages, Population and
Economic Growth, England 1209-1869”, Economic History Review, 60, 97-
135.
Clarkson, L.A. (1989), “The Manufacture of Leather”, in Mingay, G.E. (ed.), The
Agrarian History of England and Wales, Volume VI: 1750-1850, Cambridge:
Cambridge University Press, 466-483.
Crafts, N.F.R. (1976), “English Economic Growth in the Eighteenth Century: A Re-
examination of Deane and Cole’s Estimates”, Economic History Review, 29,
226-235.
Crafts, N.F.R. (1985), British Economic Growth during the Industrial Revolution,
Oxford: Oxford University Press.
Crafts, N.F.R. and Harley, C.K. (1992), “Output Growth and the Industrial
Revolution: A Restatement of the Crafts-Harley View”, Economic History
Review, 45, 703-730.
Crafts, N.F.R., Leybourne, S.J. and Mills, T.C. (1989), “Trends and Cycles in British
Industrial Production, 1700-1913”, Journal of the Royal Statistical Society,
Series A, 152, 43-60.
Deane, P. (1968), “New Estimates of Gross National Product for the United Kingdom
1830-1968”, Review of Income and Wealth, 14, 95-112.
Deane, P. and Cole, W.A. (1967), British Economic Growth, 1688-1959, Second
Edition, Cambridge: Cambridge University Press.
Feinstein, C.H. (1972), National Income, Expenditure and Output of the United
Kingdom, 1855-1965, Cambridge: Cambridge University Press.
Feinstein, C.H. (1998), “Pessimism Perpetuated: Real Wages and the Standard of
Living in Britain during and after the Industrial Revolution”, Journal of
Economic History, 58, 625-658.
Gayer, A.D., Rostow, W.W. and Schwartz, A.J. (1953), The Growth and Fluctuation
of the British Economy, 1790-1850, Oxford: Oxford University Press.
46
Harley, C. K. (1988) “Ocean Freight Rates and Productivity, 1740-1913: The Primacy
of Mechanical Invention Reaffirmed”, Journal of Economic History, 48, 851-
876.
Hodrick, R.J. and Prescott, E.C. (1997), “Postwar U.S. Business Cycles: An Empirical
Investigation”, Journal of Money, Credit and Banking, 29, 1-16.
Holderness, B.A. (1989), “Prices, productivity, and Output”, in Mingay, G.E. (ed.),
The Agrarian History of England and Wales, Volume VI: 1750-1850,
Cambridge: Cambridge University Press, 84-189.
Horrell, S., Humphries, J. and Weale, M. (1991), “Input-Output Table 1841: Notes on
Construction”, University of Cambridge.
Horrell, S., Humphries, J. and Weale, M. (1994), “An Input-Output Table for 1841”,
Economic History Review, 47, 545-566.
John, A.H. (1989), “Statistical Appendix”, in Mingay, G.E. (ed.), The Agrarian
History of England and Wales, Volume VI: 1750-1850, Cambridge:
Cambridge University Press, 972-1155.
King, P. (2005), “The Production and Consumption of Bar Iron in Early Modern
England and Wales”, Economic History Review, 58, 1-33.
Lee, C.H. (1986), The British Economy since 1700: A Macroeconomic Perspective,
Cambridge: Cambridge University Press.
Lindert, P.H. and Williamson, J.G. (1982), Revising England’s Social Tables 1688-
1913”, Explorations in Economic History, 19, 385-408.
O'Brien, P. K. (1985), “Agriculture and the Home Market for English Industry, 1660-
1820”, English Historical Review, 100, 773-800.
47
Overton, M., Whittle, J., Dean, D. and Haan, A. (2004), Production and Consumption
in English Households, 1600-1750, London: Routledge.
Perren, R. (1975), “The Meat and Livestock Trade in Britain, 1850-70”, Economic
History Review, 28, 385-400.
Prince, H.C. (1989), “The Changing Rural Landscape, 1750-1850”, in Mingay, G.E.
(ed.), The Agrarian History of England and Wales, Volume VI: 1750-1850,
Cambridge: Cambridge University Press, 7-83.
Ravn, M.O. and Uhlig, H. (2002), “On Adjusting the Hodrick-Prescott Filter for the
Frequency of Observations”, Review of Economics and Statistics, 84, 371-375.
Turner, M.E., J.V. Beckett, and B. Afton (2001), Farm Production in England 1700-
1914, Oxford and New York: Oxford University Press.
48
Wrigley, E.A., Davies, R.S., Oeppen, J.E. and Schofield, R.S. (1997), English
Population History from Family Reconstitution, 1580-1837, Cambridge:
Cambridge University Press.
Wrigley, E.A. and Schofield, R.S. (1989), The Population History of England, 1541-
1871: A Reconstruction, Cambridge: Cambridge University Press.